Interferon-signaling pathway: associations with ... - Oxford Academic

Download PDF
12 downloads 523 Views 100KB Size Report
Comment
Data were collected by trained and certified interviewers using laptop com- puters. All interviews .... version 2.11.0 (R Foundation for Statistical Computing, Vienna, Austria). This method takes .... and microsatellite instability (MSI). IFN-signaling ...
Carcinogenesis vol.32 no.11 pp.1660–1667, 2011 doi:10.1093/carcin/bgr189 Advance Access publication August 22, 2011

Interferon-signaling pathway: associations with colon and rectal cancer risk and subsequent survival Martha L.Slattery1,, Abbie Lundgreen1, Kristina L.Bondurant2 and Roger K.Wolff1 1 Department of Internal Medicine, University of Utah Health Sciences Center, 295Chipeta Way, Salt Lake City, UT 84108, USA and 2Department of Epidemiology, Fay W. Boozman College of Public Health, University of Arkansas for Medical Sciences, Little Rock, AR 72205, USA  To whom correspondence should be addressed. Tel: þ1 801 585 6955; Fax: þ1 801 581 3623; Email: [email protected]

Interferons (IFNs) are proteins involved in many functions including antiviral and antimicrobial response, apoptosis, cell cycle control and mediating other cytokines. IFN gamma (IFNG) is a proinflammatory cytokine that modulates many immune-related genes. In this study we examine genetic variation in IFNG, IFNGR1, IFNGR2 and interferon regulatory factors (IRFs) to determine associations with colon and rectal cancer and survival after diagnosis. We include data from two population-based incident studies of colon cancer (1555 cases and 1956 controls) and rectal cancer (754 cases and 959 controls). Five tagSNPs in IFNG, IRF2 and IRF3 were associated with colon cancer and eight tagSNPs in IFNGR1, IFNGR2, IRF2, IRF4, IRF6 and IRF8 were associated with rectal cancer. IRF3 rs2304204 was associated with the strongest direct association and IRF2 3775554 with the strongest inverse association for colon cancer [odds ratios (ORs) 1.43, 95% confidence interval (CI) 1.12–1.82 for recessive model and 0.52, 95% CI 0.28–0.97 for unrestricted model]. For rectal cancer, IFNGR1 rs3799488 was directly associated with risk (OR 2.30, 95% CI 1.04–5.09 for recessive model), whereas IRF6 rs861020 was inversely associated with risk (OR 0.57, 95% CI 0.34–0.95). Several single-nucleotide polymorphisms interacted significant with both NF-kB1 and IL6 and with aspirin/non-steroidal antiinflammatory drugs and cigarette smoking. Using a summary score to estimate mutational load, we observed a hazard rate ratio (HRR) close to 5.00 (95% CI 2.73–8.99) for both colon and rectal (HRR 4.83, 95% CI 2.34–10.05) cancer for those in the category having the most at-risk genotypes. These data suggest the importance of IFN-signaling pathway on colon and rectal cancer risk and survival after diagnosis.

Introduction Interferons (IFNs) are proteins involved in many functions including antiviral and antimicrobial response, apoptosis, control of cell cycle and mediators of other cytokines (1,2). There are three classes of IFNs, type I, II and III. Interferon gamma (IFNG) is the only type II IFN and as a proinflammatory cytokine, has been identified as an important modulator of immune-related genes, including nuclear factor-kappa B (NF-jB), toll-like receptor 3 (TLR3), VCAM1 and CASP4 (3), interferon gamma receptor (IFNGR), interferon regulatory factors (IRF), V-AKT murine thymoma viral oncogene homolog 1 (AKT), mitogen-activated protein kinases and inhibitor of kappa (IKK) (1, 4). IFN receptors are required for IFNs to exert their biological activity and therefore play a critical role in IFN signaling (4,5); IFNGRs have two subunits, IFNGR1 and IFNGR2. IRFs are a family of transcription factors (2,6) involved in the regulation of the Abbreviations: CI, confidence interval; HRR, hazard rate ratio; IFN, interferon; IFNG, interferon gamma; IFNGR, interferon gamma receptor; IRF, interferon regulatory factor; KPMCP, Kaiser Permanente Medical Care Program; NSAID, non-steroidal anti-inflammatory drug; OR, odds ratio.

IFN system, cell growth and the regulation of host defense such as innate and adaptive immune response. The IFN-signaling system may play a critical role in carcinogenic processes. However, few studies of genetic variation in the IFNsignaling pathway have been examined with colon or rectal cancer. Of these genes, only IFNG has been examined, perhaps because of its role in maintaining the integrity of the intestinal epithelial barrier (7). IFNG -874T . A (rs2430561) was not associated with risk of hereditary non-polyposis colon cancer in a study of 212 cases (8). A small study of 170 colon and rectal cancer cases in Korea did not find an association with IFNG 5644 (9). Studies examining genetic variation in other components of the IFNG-signaling pathway have not been reported nor have studies examined the impact of genetic variation in this pathway on survival. Given the role of IFNG in apoptosis, cell growth and regulation, such an association is biologically plausible. In this study, we examine the genetic variation in IFNG, IFNGR1, IFNGR2, IRF1, IRF2, IRF3, IRF4, IRF5, IRF6, IRF7, IRF8 and IRF9 with risk of developing colon and rectal cancer as well as their association with survival. Given the biological function of this signaling pathway, we evaluate interaction with two key inflammation-related genes, NF-jB1 and IL6 (10) as well as two lifestyle factors that may modify genetic susceptibility, use of aspirin and/or non-steroidal antiinflammatory drugs (NSAIDs) and cigarette smoking. Both aspirin/ NSAID use and cigarette smoking may modify associations through their influence on inflammation. Aspirin/NSAID use may reduce inflammation, whereas cigarette smoking may increase inflammation as a result of oxidative stress. Methods Two study populations are included. The first, a population-based case–control study of colon cancer, included cases (n 5 1555) and controls (n 5 1956) identified between 1 October 1991 and 30 September 1994 living in the Twin Cities Metropolitan Area, Kaiser Permanente Medical Care Program (KPMCP) of Northern California and a seven-county area of Utah (11). The second study used identical data collection methods as the first study but included population-based cases with cancer of the rectosigmoid junction or rectum (n 5 754) and controls (n 5 959) who were identified between May 1997 and May 2001 in Utah and KPMCP (12). Eligible cases were between 30 and 79 years old at time of diagnosis, English speaking, mentally competent to complete the interview, no previous history of CRC and no known (as indicated on the pathology report) familial adenomatous polyposis, ulcerative colitis or Crohn’s disease. Controls were matched to cases by sex and by 5 years age groups. At KPMCP, controls were randomly selected from membership lists. In Utah, controls 65 years were randomly selected from the Health Care Financing Administration lists and controls ,65 years were randomly selected from driver’s license lists. While in Minnesota, controls were selected from driver’s license and state-identification lists. Study details have been previously reported (11,12). Interview data collection Data were collected by trained and certified interviewers using laptop computers. All interviews were audiotaped and reviewed for quality control purposes (13). The referent period for the study was 2 years prior to diagnosis for cases and prior to selection for controls. Detailed information was collected on diet, physical activity, medical history and cigarette smoking history, regular use of aspirin and NSAIDs and body size. Regular use of aspirin/NSAIDs was defined as at least three times a week for at least 1 month. Tumor registry data Tumor registry data were obtained to determine disease stage at diagnosis and months of survival after diagnosis. Disease stage was categorized centrally by one pathologist in Utah using the sixth edition of the American Joint Committee on Cancer (AJCC) staging criteria. Local tumor registries also provided information on patient follow-up including vital status, cause of death and contributing cause of death. Follow-up was obtained for all study participants for at least 5 years and was terminated for the Colon Cancer Study in 2000 and

Ó The Author 2011. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

1660

IFN-signaling pathway

for the Rectal Cancer Study in 2007. We used the standard definition of colon and rectal cancer employeed by the Surveillance and Epidemiology and End Results (SEER) program. TagSNP selection and genotyping TagSNPs were selected using the following parameters: linkage disequilibrium blocks were defined using a Caucasian linkage disequilibrium map and an r2 5 0.8; minor allele frequency .0.1; range 5 1500 bps from the initiation codon to þ1500 bps from the termination codon and one single-nucleotide polymorphism (SNP)/linkage disequilibrium bin. All markers were genotyped using a multiplexed bead-array assay format based on Golden Gate chemistry (Illumina, San Diego, CA). A genotyping call rate of 99.85% was attained. Blinded internal replicates represented 4.4% of the sample set; the duplicate concordance rate was 100%. Individuals with missing genotype data were not included in the analysis for that specific marker. We evaluated associations with 12 candidate genes, including IFNG (3 SNPs), IRNGR1 (4 SNPs), IFNGR2 (5 SNPs), IRF1 (2 SNPs), IRF2 (51 SNPs), IRF3 (2 SNPs), IRF4 (10 SNPs), IRF5 (4 SNPs), IRF6 (5 SNPs), IRF7 (2 SNPs), IRF8 (12 SNPs) and IRF9 (2 SNPs). Tumor marker data We have previously evaluated tumors for CpG island methylator phenotype, microsatellite instability, TP53 mutations and KRAS2 mutations (14–17) and were therefore able to evaluate genes in relation to tumors with specific characteristics or markers. Details for methods used to evaluate these epigenetic and genetic changes have been described in previous publications (14–17). Statistical methods Statistical analyses were performed using SASÒ version 9.2 (SAS Institute, Cary, NC). We report odds ratios (ORs) and 95% confidence intervals (95% CIs) assessed from multiple logistic regression models adjusting for age, center, race/ethnicity and sex. To summarize risk associated with multiple variants across the pathway, we created a summary polygenic score that was based on all at-risk genotypes for colon and rectal cancer. The score for each SNP was based on the inheritance model and its associated risk. For the codominant or additive model, a score of 0, 1 or 2 was assigned directly correlated to the number of high-risk alleles; scores of 0 or 2 were assigned for the dominant and recessive models. After assigning a score for each SNP previously identified as being significant, the scores were summed across SNPs to generate an individual polygenic summary score. Individuals missing SNP data were dropped from the analysis. The continuous score variable was redefined as a categorical variable based on the frequency distribution within the study population. Analysis for interaction was based on tagSNPs within each gene. Lifestyle variables were selected because of their biological plausibility for involvement in this candidate pathway; in these analyses, we focused on interaction between cigarette smoking and use of aspirin/NSAIDs. We tested interaction with two genes, NF-jB1 and IL6, which we hypothesized as importantly modifying the effect of candidate genes being analyzed given their importance in inflammatory processes. P values for interaction were determined using a likelihoodratio test comparing a full model that included an ordinal interaction term with a reduced model without an interaction term. Survival months were calculated based on month and year of diagnosis, and month and year of death or date of last contact. Associations between SNPs and risk of dying of colorectal cancer were evaluated using Cox proportional hazards models to obtain multivariate hazard rate ratios (HRRs) and 95% CIs. We adjusted for age at diagnosis, study center, race, sex, tumor molecular phenotype and AJCC stage to estimate HRRs. Adjusted multiple comparison P values, taking into account tagSNPs within the gene, were estimated using the methods by Conneely et al. (18) via R version 2.11.0 (R Foundation for Statistical Computing, Vienna, Austria). This method takes into account the correlated nature of the SNP data within a gene. Wald P values from the main effect models and interaction P values based on likelihood-ratio tests were used for estimates of multiple comparisons. We consider a pACT of ,0.20 as being potentially important given the candidate pathway approach and the need to consider both type 1 and type 2 errors. We believe that findings at this level would merit replication, especially when evaluating interactions.

Results The population characteristics are described in Table I. The colon cancer study consisted of cases and controls from all the three centers, whereas the rectal cancer study only included cases and controls for KPMCP and Utah. The majority of the population was non-Hispanic white, male and .60 years of age. The genes with corresponding

Table I. Description of study population Colon Control n (%)

Rectal Case n (%)

Total 1956 1555 Age 30–39 40 (2.04) 23 (1.48) 40–49 128 (6.54) 102 (6.56) 50–59 326 (16.67) 290 (18.65) 60–69 673 (34.41) 538 (34.60) 70–79 789 (40.34) 602 (38.71) Center Utah 378 (19.33) 249 (16.01) KPMCP 787 (40.24) 744 (47.85) Minnesota 791 (40.44) 562 (36.14) Race/ethnicity NHW 1828 (93.46) 1428 (91.83) Hispanics 75 (3.83) 59 (3.79) Black 53 (2.71) 68 (4.37) Asian 0 0 Sex Male 1047 (53.53) 870 (55.95) Female 909 (46.47) 685 (44.05) AJCC stage Stage I 469 (30.16) Stage II 405 (26.05) Stage III 374 (24.05) Stage IV 128 (8.23) Unknown 179 (11.51) Tumor molecular phenotypes KRAS2 mutation 348 (22.38) TP53 mutation 516 (33.18) 272 (17.49) CIMPa high 185 (11.90) MSIa unstable

Control n (%)

Case n (%)

959

754

21 (2.19) 19 (2.52) 101 (10.53) 96 (12.73) 243 (25.34) 196 (25.99) 329 (34.31) 250 (33.16) 265 (27.63) 193 (25.60) 365 (38.06) 274 (36.34) 594 (61.94) 480 (63.66) 0 0 824 (85.92) 625 (82.89) 63 (6.57) 61 (8.09) 43 (4.48) 29 (3.85) 29 (3.02) 39 (5.17) 541 (56.41) 451 (59.81) 418 (43.59) 303 (40.19) 381 (50.53) 124 (16.45) 175 (23.21) 57 (7.56) 17 (2.25) 173 (22.94) 277 (36.74) 59 (7.82) 14 (1.86)

a

Tumor molecular phenotypes are CpG island methylator phenotype (CIMP) and microsatellite instability (MSI).

tagSNPs that were associated with either colon or rectal cancer independently or through interaction with gene or lifestyle factors are described in Table II. All SNPs were in Hardy–Weinberg equilibrium. Roughly 90% of the population was non-Hispanic white. A summary of all SNPs analyzed can be found in the Supplementary Table, available at Carcinogenesis Online. Five tagSNPs in three genes (IFNG, IRF2 and IRF3) were associated with colon cancer (Table III) and eight tagSNPs in six genes (IFNGR1, IFNGR2, IRF2, IRF4, IRF6 and IRF8) were associated with rectal cancer. The strongest increased risk was associated with IRF3 rs2304204 for colon cancer (OR 1.43, 95% CI 1.12–1.82 for recessive model) and the strongest inverse association was observed for IRF2 rs3775554 (OR 0.52, 95% CI 0.28–0.97 for unrestricted or codominant model or unrestricted). For rectal, rs3799488 of IFNGR1 was associated with over a 2-fold increased risk (OR 2.30, 95 % CI 1.04–5.09 for recessive model), whereas IRF6 rs861020 was associated with the strongest inverse association (OR 0.57, 95% CI 0.34– 0.95). Only two SNPs in IRF2 were associated with colorectal cancer when colon and rectal cancer were combined. The risk estimate for IRF2 rs3733473 for colon cancer was 0.63 (95% CI 0.43–0.92), for rectal cancer was 0.97 (95% CI 0.61–1.53) and for the colorectal cancer was 0.74 (95% CI 0.55–0.99) with the association clearly being driven by colon cancer. On the other hand, a trend toward a protective effect of IRF2 rs7655800 was seen for both colon and rectal cancer (OR 0.62, 95% CI 0.39–1.00 for colon cancer; OR 0.75, 95% CI 0.36–1.59 for rectal cancer; OR 0.66, 95% CI 0.44–0.98 for colorectal cancer). The Supplementary Table, available at Carcinogenesis Online, shows risk associated with all SNPs for colorectal cancer. Genes in this pathway appeared to be most uniquely associated with CpG island methylator phenotype þ tumors (P for heterogeneity ,0.05 for IRF2 rs3733473, rs6812958 and IRF6 rs17015218 for

1661

M.L.Slattery et al.

Table II. Descriptive table of tagSNPs associated with colon and rectal cancer Symbol

Chromosome location

Alias

SNP

Major/minor allele

MAFa

FDR HWEb

IFNG

12q14

IFG IFI

IFNGR1

6q23–q24

IFNGR2

21q22.11

CD119 FLJ45734 IFNGR AF-1 IFGR2 IFNGT1

IRF1

5q31.1

rs1861493 rs2069718 rs2069727 rs1327474 rs1327475 rs3799488 rs1532 rs2834211 rs2834213 rs2834215 rs17622656

A/G C/T A/G A/G C/T T/C C/T T/C A/G G/A G/A

0.31 0.42 0.46 0.46 0.16 0.12 0.30 0.10 0.24 0.45 0.37

1.00 1.00 1.00 1.00 1.00 1.00 0.95 0.76 0.96 0.87 1.00

IRF2

4q34.1–q35.1

IRF3 IRF4

19q13.3–q13.4 6p25–p23

IRF5

7q32

IRF6

1q32.3–q41

rs793801 rs809909 rs965225 rs2797507 rs3733473 rs3756093 rs3756094 rs3775554 rs3775556 rs3775574 rs3822118 rs6812958 rs6827018 rs6856910 rs7655800 rs9684244 rs11132242 rs12512614 rs17488206 rs17585389 rs2304204 rs872071 rs1050975 rs11242865 rs3778607 rs3800262 rs7768807 rs12211228 rs752637 rs1874328 rs861020 rs2013162 rs2013196 rs17015218

G/A T/A G/A C/A G/A C/G G/A G/C A/G A/G C/T G/A A/G T/C A/G G/C A/G G/T A/T T/C A/G A/G A/G C/T G/A G/A T/C G/C G/A T/C G/A C/A C/T A/G

0.40 0.33 0.09 0.46 0.20 0.15 0.32 0.13 0.26 0.40 0.32 0.28 0.14 0.33 0.15 0.37 0.38 0.24 0.24 0.27 0.26 0.50 0.09 0.20 0.48 0.17 0.25 0.14 0.37 0.37 0.21 0.38 0.20 0.16

0.22 0.96 1.00 1.00 1.00 0.61 1.00 1.00 0.95 0.92 0.96 0.79 0.93 1.00 1.00 1.00 0.97 0.68 0.98 1.00 0.95 1.00 0.94 0.94 1.00 1.00 1.00 1.00 0.89 0.92 1.00 1.00 0.68 0.97

IRF7

11p15.5

rs1131665

A/G

0.26

1.00

IRF8

16q24.1

rs305084 rs1044873 rs305071 rs13338943

T/C C/T G/A G/T

0.09 0.39 0.12 0.11

0.97 0.96 0.96 1.00

IRF-1 MAR DKFZp686F0244 IRF2

LSIRF MUM1

LPS OFC6 PIT PPS VWS IRF-7H IRF7A H-ICSBP ICSBP ICSBP1 IRF8

a

Minor allele frequency (MAF) based on control for non-Hispanic white population. FDR (HWE), false discovery rate adjusted P value for Hardy–Weinberg equilibrium test; HWE based on NHW control population (sample sizes range from 2453 to 2652).

b

colon cancer and IFNG rs2069718, IRF2 rs2310047 and rs7657540 for rectal cancer) and KRAS-mutated tumors (P for heterogeneity ,0.05 for IRF6 rs2013162 for colon cancer and IRF2 rs3775556 and IRF8 rs8064189 for rectal cancer) (data not shown in table). Results were similar when analysis excluding non-Hispanic white individuals was performed. We evaluate interactions between our candidate genes and NF-jB1 and IL6, two genes we hypothesize as interacting with IFN-related genes given their role in inflammation. We have previously reported independent associations between NF-jB1 and IL6 and colon and rectal cancer (10,19) IL6 rs2069860 was associated with reduced risk of colon cancer (adjusted OR 0.55, 95% CI 0.32–0.95). NF-jB1 was

1662

associated with reduced risk of colon cancer (rs4648110 OR 0.66, 95% CI 0.45–0.96 for recessive model and rs13117745 OR 0.64, 95% CI 0.39–1.04 for recessive model). NF-jB1 also was associated with rectal cancer (OR 0.79, 95% CI 0.51–0.94 for dominant model of rs23051; OR 1.32, 95% CI 1.00–1.75 for additive model of rs3821958; OR 1.24, 95% CI 1.03–1.51 for dominant model of rs11722146). We observe numerous interactions (Table IV). For colon cancer, IFNG, IRF1, IRF2 interacted with NF-jB1, whereas IFNGR1, IFNGR2, IRF2, IRF5, IRF6 and IRF8 all interacted with IL6. For rectal cancer, we observed significant interactions between NF-jB1 and IFNG, IFNGR2, IRF4 and IRF6, whereas IRFNGR1, IFNGR2, IRF1, IRF2 and IRF8 interacted with IL6.

IFN-signaling pathway

Table III. Associations between candidate SNPs and colon and rectal cancer Controls N Colon IFNG (rs1861493) AA AG/GG IRF2 rs3775554 (G . C) GG GC CC IRF2 (rs793801) GG GA/AA IRF2 (rs809909) TT TA/AA IRF3 (rs2304204) AA/AG GG Summary score (0–2) (3–5) (6–7) (8–10) Ptrend Rectal IFNGR1 (rs3799488) TT/TC CC IFNGR2 (rs2834211) TT TC/CC IRF2 (rs3733473) GG GA/AA IRF2 (rs3775556) AA AG GG IRF2 (rs3775574) AA AG/GG IRF4 (rs11242865) CC/CT TT IRF6 (rs861020) GG/GA AA IRF8 (rs1044873) CC/CT TT Summary score (0–4) (5–6) (7–8) (9–12)

Cases N

Wald a

OR (95% CI)

947 1008

849 704

1.00 0.79 (0.69–0.91)

1479 443 34

1218 322 15

1.00 0.88 (0.74–1.03) 0.52 (0.28–0.97)

666 1290

590 964

1.00 0.84 (0.73–0.96)

878 1078

626 929

1.00 1.20 (1.04–1.37)

1817 139

1397 157

1.00 1.43 (1.12–1.82)

305 911 555 185 ,0.0001

172 653 524 206

1.00 1.27 (1.02–1.57) 1.66 (1.33–2.08) 1.92 (1.46–2.53)

949 10

737 17

1.00 2.30 (1.04–5.09)

764 195

565 189

1.00 1.31 (1.04–1.64)

625 334

449 305

1.00 1.27 (1.04–1.55)

532 370 55

369 328 57

1 1.26 (1.03–1.54) 1.48 (0.99–2.19)

317 642

297 457

1.00 0.74 (0.60–0.90)

935 24

720 34

1.00 1.90 (1.11–3.24)

910 49

732 22

1.00 0.57 (0.34–0.95)

789 170

654 100

1.00 0.71 (0.55–0.93)

251 390 229 89

122 280 232 120

1.00 1.47 (1.13–1.92) 2.10 (1.58–2.79) 2.79 (1.96–3.96)

P value

pACT

0.0007

0.0019

0.0205

0.5099

0.0112

0.3392

0.0097

0.3093

0.0041

0.0080

0.0400

0.1120

0.0222

0.0949

0.0173

0.4518

0.0071

0.2338

0.0029

0.1119

0.0184

0.1458

0.0313

0.1296

0.0142

0.1317

a

OR and 95% CI adjusted for age, center, race/ethnicity and sex.

Significant interactions between recent regular use of aspirin/ NSAIDs and smoking cigarettes with candidate genes are shown in Table V. Several genes interacted with aspirin/NSAIDs, including IRF2, IRF4, IRF5 and IRF6 for colon cancer and IFNGR2, IRF2, IRF6 and IRF7 for rectal cancer. Likewise, we observed several significant interactions between cigarette smoking and candidate genes. IRF2 and IRF4 interacted with smoking for both colon and rectal cancer. Additionally, there was significant interaction between IRF6 and smoking for colon cancer and IRF8 and smoking for rectal cancer. We evaluate pathway tagSNPs with survival by looking at the mutational load using a summary score consisting of those SNPs associated with survival based on significant HRRs (Table VI). For colon

cancer, the HRR was 4.96 (95% CI 2.73–8.99) for those in the category having the most at-risk genotypes; for rectal cancer, the upper summary HRR was 4.85 (95% CI 2.34–10.05) after adjusting for age, center, race, sex, AJCC stage and tumor molecular phenotype. Assessment of rectosigmoid junction separate from other rectal tumors showed similar results as for the combined group. Discussion Our data support the hypothesis that genetic variation in the IFNG, its receptors and IRF genes are associated with risk of developing colon and rectal cancer and that this association may be modified by other

1663

M.L.Slattery et al.

Table IV. Associations between IFNG, IRF genes and IL6, NFKB1 relative to wild-type/wild-type IFNG, IRF gene

SNP (model)a

Pathway gene

SNP (model)

Wild-typeb variant OR (95% CI)

Variantc wild-type OR (95% CI)

Variantd variant OR (95% CI)

Interaction P value

pACT

Colon IFNG

rs2069718 (A)

NFKB1

rs2069727 (A)

NFKB1

rs1327475 (D) rs2834213 (A) rs17622656 (A) rs2797507 (A) rs3756094 (A) rs6812958 (A) rs752637 (A) rs17015218 (D) rs305071 (D)

IL6

rs13117745 (D) rs4648090 (D) rs4648110 (D) rs11722146 (A) rs4648110 (D) rs4648127 (D) rs1800797 (A)

1.27 (0.98–1.63) 1.12 (0.86–1.45) 1.26 (1.00–1.60) 1.28 (0.85–1.94) 0.75 (0.59–0.97) 0.83 (0.56–1.22) 0.78 (0.62–0.99) 0.98 (0.76–1.28) 0.86 (0.63–1.16) 0.51 (0.35–0.76) 1.19 (0.95–1.49) 0.64 (0.48–0.85) 0.72 (0.51–1.01) 0.75 (0.58–0.98) 0.75 (0.61–0.92)

1.02 (0.81–1.29) 0.95 (0.76–1.19) 1.08 (0.85–1.37) 1.44 (1.10–1.88) 0.91 (0.72–1.15) 1.07 (0.87–1.31) 0.88 (0.69–1.12) 1.55 (0.93–2.59) 0.64 (0.43–0.93) 0.77 (0.57–1.04) 1.39 (1.07–1.81) 0.52 (0.34–0.81) 0.76 (0.55–1.04) 0.83 (0.66–1.03) 0.94 (0.79–1.11)

0.64 (0.46–0.89) 0.63 (0.44–0.90) 0.68 (0.50–0.93) 1.19 (0.73–1.95) 1.32 (0.99–1.75) 1.63 (1.05–2.54) 1.21 (0.86–1.71) 0.36 (0.17–0.77) 1.00 (0.65–1.53) 0.96 (0.62–1.47) 0.74 (0.49–1.12) 0.75 (0.39–1.43) 0.97 (0.57–1.64) 1.05 (0.73–1.51) 1.20 (0.86–1.67)

0.0015 0.0247 0.0008 0.0367 0.0014 0.0453 0.0142 0.0209 0.0447 0.0039 0.0010 0.0014 0.0227 0.0200 0.0106

0.0098 0.1174 0.0056 0.1611 0.0094 0.1877 0.1070 0.1851 0.1451 0.2651 0.1115 0.1183 0.1083 0.1568 0.1762

rs1861493 (A) rs1327474 (A) rs1532 (A)

NFKB1 IL6 NFKB1

rs2834215 (A)

IL6

0.78 (0.49–1.24) 1.83 (1.21–2.79) 1.87 (1.29–2.71) 0.91 (0.61–1.36) 0.86 (0.50–1.48) 1.17 (0.64–2.11) 0.58 (0.35–0.96) 0.98 (0.69–1.40) 0.61 (0.39–0.94) 2.09 (1.28–3.41) 0.85 (0.59–1.21) 1.31 (1.03–1.67) 1.33 (1.02–1.73) 0.89 (0.62–1.27) 0.86 (0.65–1.12) 0.89 (0.70–1.13) 1.54 (1.14–2.08)

1.15 (0.79–1.67) 0.94 (0.69–1.28) 1.25 (0.85–1.84) 1.66 (0.87–3.16) 0.90 (0.66–1.22) 1.79 (1.17–2.72) 0.96 (0.56–1.66) 1.01 (0.65–1.59) 0.56 (0.29–1.08) 1.19 (0.88–1.59) 1.25 (0.89–1.76) 1.43 (1.10–1.84) 1.66 (1.02–2.70) 1.47 (1.03–2.08) 0.98 (0.77–1.24) 0.93 (0.72–1.20) 1.07 (0.83–1.38)

3.33 (1.03–10.74) 0.76 (0.39–1.48) 0.35 (0.11–1.06) 1.06 (0.46–2.46) 2.31 (1.38–3.85) 1.41 (0.76–2.63) 1.31 (0.65–2.66) 4.38 (1.19–16.06) 0.86 (0.35–2.13) 0.70 (0.37–1.33) 0.52 (0.32–0.85) 1.01 (0.74–1.39) 1.06 (0.62–1.81) 0.59 (0.37–0.93) 1.62 (1.13–2.33) 1.52 (1.10–2.10) 0.67 (0.36–1.25)

0.0097 0.0338 0.0006 0.0349 0.0026 0.0337 0.0382 0.0032 0.0034 0.0021 0.0142 0.0046 0.0132 0.0146 0.0056 0.0053 0.0107

0.0636 0.2140 0.0069 0.2953 0.0267 0.2466 0.1131 0.2299 0.2375 0.1622 0.2432 0.0994 0.2314 0.1451 0.0661 0.0621 0.1885

IFNGR1 IFNGR2 IRF1 IRF2 IRF5 IRF6 IRF8 Rectal IFNG IFNGR1 IFNGR2

IRF1 IRF2

IRF6

rs17622656 (A) rs12512614 (A) rs17585389 (A) rs2797507 (A) rs11242865 (D) rs3800262 (D) rs7768807 (A) rs2013196 (D)

IRF8

rs13338943 (D)

IRF4

NFKB1 IL6 NFKB1 IL6

rs230510 (A) rs1800797 (A) rs4648090 (D) rs1800797 (A) rs2069840 (A) rs2069827 (D)

IL6 NFKB1

IL6

rs4648127 (D) rs1800796 (D) rs230510 (A) rs1800796 (D) rs2069840 (A) rs1800797 (A) rs1800796 (D) rs2069840 (A) rs1800796 (D) rs230510 (A) rs4648110 (D) rs4648110 (D) rs230510 (A) rs4648090 (D) rs4648110 (D) rs1800796 (D)

a

Models: A, additive or codominant; D, dominant. Compares wild-type (WT) IFNG/IRF gene and variant from additive model or heterozygote/variant if dominant model for pathway SNP relative to both WT. c Compares variant from additive model or heterozygote/variant if dominant model for IFGN/IRF gene and WT pathway gene relative to both WT. d Compares variant from additive model or heterozygote/variant if dominant model for both IFNG/IRF and pathway gene relative to both WT. b

key inflammation-related genes and lifestyle factors such as use of aspirin/NSAID and cigarette smoking. Additionally, we provide support for the hypothesis that genetic variation in the IFNG-signaling pathway is associated with survival. The increased risk of both developing colon or rectal cancer and survival after diagnosis appears to be influenced by mutational load. Our data suggest that unique associations were observed for CpG island methylator phenotype þ and KRAS2-mutated tumors, suggesting that these tumor molecular phenotypes may be associated with inflammation. IRF2 was associated with both colon and rectal cancer, whereas other components of the pathway were uniquely associated with colon cancer, (i.e. IFNG and IRF3), and with rectal cancer (i.e. IFNGR1, IFNGR2, IRF4, IRF6 and IRF8). Although we acknowledge that these differences could stem from chance findings, many associations remained significant after adjusting for multiple comparisons. These findings also could support other reports showing differences in both genetic and lifestyle factors for colon and rectal cancer (12,20–23). For instance, body size and insulin signaling may play a larger role in the etiology of colon versus rectal cancer (21,23,24). Studies have shown that IFNG attenuates insulin signaling (25); thus an association between IFNG and colon cancer may reflect different biological components of colon versus rectal cancer. Inflammation is a key element in colon and rectal carcinogenesis. We evaluated the interaction of IFN-signaling pathway genes with NF-jB1 and IL6, two genes that appear to be pivotal in inflammatory response. All genes, except IRF3, IRF4 and IRF9 for colon cancer, and IRF3, IRF5, IRF7 and IRF9 for rectal cancer showed significant interaction with these genes. Others have shown that NF-jB1

1664

expression is influenced by the IFN-signaling pathway (3). We interpret these findings to indicate both the importance of IFN-signaling pathway to an inflammation-related mechanism as well as the degree to which multiple inflammation factors work together to influence cancer risk. Although beyond the scope of this paper, we believe that it is important to examine how IFN genes work with other inflammation-related genes given the number of interactions observed. Genes that may be important include tumor necrosis factor and its receptors, toll-like receptors, mitogen-activated protein kinases including p38, mitogenactivated protein kinase 8 and mitogen-activated protein 14, inhibitor of kappa light chain gene enhancer in B cells, kinase of Beta (IKKB), cytokines such as interleukin 1 and interleukin 8 and AKT in addition to angiogenesis genes such as vascular endothelial growth factor and its receptors. Our data suggest that genetic susceptibility is influenced by regular use of aspirin/NSAID use and smoking cigarettes. The role of aspirin and NSAID use in colon and rectal cancer risk are well documented (26–29). These associations are felt to stem from the anti-inflammatory properties of these drugs. Cigarette smoking has been associated with increased nitric oxide (NO) synthesis by activating nitric oxide synthase (NOS2) (30,31); NO has been shown to contribute to chronic inflammation (32). While multiple genes were associated with both colon and rectal cancer, IRF2 and IRF6 were associated with aspirin/ NSAID use for both colon and rectal cancer and IRF2 was associated with cigarette smoking for both colon and rectal cancer. Few studies have examined how either aspirin/NSAID or cigarette smoking works with these genes, although the interaction with genes in the IFNsignaling pathway is biologically plausible. One study has shown that

IFN-signaling pathway

Table V. Interaction between cigarette smoking, NSAID use and IFNGR2, IRF genes and risk of colon and rectal cancer relative to wild-type/wild-type IFNG, IRF genes

Colon IRF2

IRF4 IRF5 IRF6 Rectal IFNGR2 IRF2

IRF6 IRF7 Colon IRF2 IRF4 IRF6 Rectal IRF2

IRF4 IRF8

SNP (model)a

Variantb

Wild-type

Interaction P value

pACT

No regular aspirin/NSAID use OR (95% CI)

Regular aspirin/NSAID use OR (95% CI)

Regular aspirin/NSAID use OR (95% CI)

rs3756093 (D) rs3822118 (A) rs6856910 (A) rs9684244 (A) rs1050975 (D) rs3778607 (A) rs1874328 (A) rs2013162 (A) rs2013196 (D)

1.17 (0.97–1.41) 0.84 (0.62–1.13) 1.24 (0.94–1.65) 1.23 (0.94–1.60) 0.83 (0.67–1.03) 0.90 (0.71–1.14) 1.01 (0.77–1.31) 0.72 (0.56–0.94) 0.80 (0.67–0.96)

0.56 (0.44–0.71) 0.84 (0.56–1.24) 0.52 (0.36–0.76) 0.65 (0.46–0.90) 0.76 (0.58–1.01) 0.73 (0.56–0.97) 0.77 (0.54–1.10) 0.64 (0.46–0.88) 0.64 (0.52–0.80)

0.74 (0.62–0.87) 0.58 (0.47–0.71) 0.79 (0.64–0.98) 0.85 (0.68–1.07) 0.60 (0.51–0.70) 0.54 (0.41–0.72) 0.53 (0.42–0.67) 0.53 (0.42–0.67) 0.59 (0.49–0.70)

0.0070 0.0349 0.0047 0.0056 0.0189 0.0454. 0.0338 0.0136 0.0410

0.2194 0.6535 0.1631 0.1871 0.1411 0.2766 0.0960 0.0600 0.1525

rs1532 (A) rs3756093 (D) rs3775574 (D) rs3822118 (A) rs965225 (D) rs861020 (D) rs1131665 (A)

0.75 (0.47–1.20) 1.08 (0.82–1.43) 0.87 (0.67–1.14) 0.63 (0.41–0.98) 1.11 (0.80–1.55) 1.17 (0.90–1.52) 1.91 (1.15–3.16) Non-recent smoker

0.82 (0.47–1.44) 0.54 (0.38–0.76) 0.51 (0.38–0.68) 1.18 (0.69–2.00) 0.48 (0.32–0.74) 0.61 (0.45–0.83) 0.55 (0.31–0.97) Recent smoker

0.54 (0.41–0.71) 0.79 (0.63–1.00) 0.92 (0.66–1.26) 0.51 (0.38–0.69) 0.76 (0.61–0.94) 0.82 (0.64–1.05) 0.82 (0.63–1.07)

0.0094 0.0415 0.0303 0.0004 0.0451 0.0289 0.0141

0.0438 0.7038 0.6056 0.0162 0.7242 0.1234 0.0280

rs6827018 (D) rs11242865 (D) rs12211228 (D) rs872071 (A) rs861020 (D)

1.12 (0.95–1.32) 0.92 (0.79–1.08) 0.90 (0.75–1.06) 1.21 (0.98–1.50) 1.13 (0.97–1.32)

0.97 (0.71–1.33) 1.51 (1.15–1.98) 1.59 (1.17–2.18) 1.04 (0.73–1.48) 1.02 (0.78–1.34)

1.29 (1.06–1.57) 0.96 (0.77–1.19) 1.00 (0.82–1.22) 1.93 (1.39–2.68) 1.35 (1.09–1.68)

0.0442 0.0030 0.0035 0.0007 0.0237

0.7598 0.0239 0.0256 0.0062 0.1011

rs11132242 (A) rs12512614 (A) rs17488206 (D) rs3756094 (A) rs7655800 (D) rs3800262 (D) rs7768807 (A) rs305084 (D)

1.33 (0.96–1.84) 1.42 (0.91–2.21) 0.83 (0.67–1.03) 1.14 (0.78–1.66) 0.98 (0.78–1.25) 1.27 (1.02–1.59) 1.56 (1.05–2.31) 0.83 (0.62–1.11)

0.7 (0.40–1.48) 1.01 (0.35–2.95) 1.69 (1.13–2.52) 0.53 (0.24–1.18) 1.94 (1.22–3.06) 1.09 (0.71–1.67) 0.55 (0.21–1.44) 2.26 (1.18–4.32)

1.69 (1.14–2.49) 1.64 (1.18–2.28) 0.97 (0.70–1.34) 1.56 (1.08–2.28) 1.10 (0.81–1.48) 1.57 (1.16–2.13) 1.61 (1.16–2.24) 1.14 (0.87–1.50)

0.0118 0.0214 0.0053 0.0272 0.0428 0.0269 0.0063 0.0198

0.3415 0.5131 0.1862 0.5866 0.7340 0.1848 0.0530 0.1654

a

Models: A, additive or codominant; D, dominant. Heterozygote/variant genotype if dominant model, variant if recessive; all comparisons are made to non-user/smoker and wild-type.

b

the NOS2 promoter contained sequences for several transcription factors including IRF6; exposure to tobacco smoke caused IRF6 to bind to the NOS2 promoter regulating NOS2 transcription and the cell response to tobacco exposure (31). Other studies like the one by Ratovitski will provide additional insight into the functionality of these genes. Functions of the IFN-signaling pathway include apoptosis and cell proliferation. IRF1 has been shown to play a role in suppression of growth of breast cancer cells (33). IFNG has been shown to regulate the expression of apoptosis-related genes and has been hypothesized to regulate cell sensitivity to apoptosis (34). Additionally, IFNG has been shown to work with tumor necrosis factor to overcome resistance of metastatic colon tumor cells to the tumor necrosis factor-related apoptosis-inducing ligand, which is an immune effector molecule (35). Our observations that genetic variations in the IFN-signaling pathway influence survival have merit. The observed risk associated with multiple variants within the pathway further suggests that the mutational load is important. With increasing number of variant genotypes, the risk of dying increased. Although one could hypothesize that a single insult to the pathway could influence risk and that additional insults would have minimal effect on risk, our data suggest otherwise. Inflammatory pathways are somewhat redundant, composed of multiple cytokines with overlapping functions; this supports that multiple insults to the pathways would result in increased risk. Our data support the hypothesis that increases in risk and hazard of dying is linear and that as mutational load of high-risk genotypes increase, so does the risk of developing cancer and dying after being diagnosed with cancer. Our observed

increased risk of dying was independent of disease stage at diagnosis and tumor molecular phenotype. Major strengths of our study were the hypothesis-driven approach, the large and extensive data set includes information on genetic, lifestyle, tumor and survival data, and our ability to examine colon and rectal cancer separately. Although we believe that the data we present is both thorough and informative, we acknowledge that limitations exist. For instance, while we have detected associations, we have minimal information on the functionality of SNPs evaluated. Additional labbased work is needed to determine functionality. We have limited our assessment of interaction to NF-jB and IL6, although other genes such as TLR3, VCAM3 and CASP4 were not considered. Additionally, we have made many comparisons. We have provided pACT values, which account for these comparisons although chance findings may exist. A hazard of multiple testing adjustments is the increased likelihood of rejecting a finding that is true. Thus, we believe that adjusted P values of ,0.20, especially for interactions, merit replication in other large sample sets to validate these findings. Our assessment was limited to those enrolled in the study; those with the poorest survival were less likely to be included which, however, we did not observed differences in association when we examined disease stage at time of diagnosis. We conclude that genetic variation in the IFN-signaling pathway is important in the etiology of colon and rectal cancer. These associations appear to be modified by lifestyle factors such as aspirin/NSAID use and cigarette smoking and other inflammation-related genes. Additionally, our data suggest the importance of genetic variation in this pathway on survival after diagnosis. We encourage validation of these findings in other large studies.

1665

M.L.Slattery et al.

Table VI. Association between survival and IFN-related genes adjusted for age, center, race, sex, AJCC stage and tumor molecular phenotype

Colon IFNGR1 (rs1327474) AA/AG GG IFNGR1 (rs9376267) CC CT/TT IFNGR2 (rs2834211) TT TC/CC IRF2 (rs12504466) TT TC/CC IRF2 (rs13116389) GG GT/TT IRF2 (rs2797507) CC CA/AA IRF2 (rs3775582) GG GA/AA IRF2 (rs7655800) AA AG/GG IRF2 (rs793777) CC CG GG IRF2 (rs793801) GG/GA AA IRF2 (rs793814) TT/TA AA IRF2 (rs9684244) GG/GC CC IRF6 (rs2013196) CC CT/TT IRF8 (rs1044873) CC CT/TT IRF8 (rs305083) AA AG/GG Summary score (2–10) (11–12) (13–14) (15–16) (17–18) (19–20) (21–22) (23–28) Ptrend Rectal IFNGR2 (rs2834213) AA/AG GG IRF2 (rs1425551) AA/AC CC IRF2 (rs3756094) GG/GA AA IRF2 (rs3822118) CC CT/TT

1666

Death/person-years

HRR (95% CI)

261/6417 48/1731

1.00 0.69 (0.50–0.94)

181/5109 128/3039

1.00 1.37 (1.09–1.73)

233/6468 76/1680

1.00 1.32 (1.01–1.72)

67/2349 242/5799

1.00 1.51 (1.14–1.99)

190/5662 119/2486

1.00 1.38 (1.09–1.75)

107/2528 202/5620

1.00 0.77 (0.61–0.98)

248/6222 61/1926

1.00 0.67 (0.50–0.89)

212/6142 97/2006

1.00 1.33 (1.04–1.70)

134/3058 141/3829 34/1260

1.00 0.89 (0.70–1.14) 0.67 (0.46–0.98)

262/6911 47/1232

1.00 1.39 (1.01–1.91)

277/7120 32/1015

1.00 0.57 (0.39–0.83)

282/6923 27/1225

1.00 0.58 (0.39–0.87)

190/5451 119/2681

1.00 1.29 (1.02–1.63)

101/3073 208/5075

1.00 1.32 (1.04–1.68)

184/5181 125/2967

1.00 1.31 (1.04–1.65)

26/1204 32/1018 45/1440 55/1465 50/1381 46/851 35/497 20/293

1.00 2.06 (1.21–3.49) 2.13 (1.31–3.48) 2.03 (1.27–3.26) 3.00 (1.85–4.87) 3.43 (2.10–5.60) 4.26 (2.53–7.19) 4.96 (2.73–8.99) ,0.0001

155/4084 16/205

1.00 2.04 (1.16–3.57)

133/3334 38/956

1.00 1.50 (1.03–2.18)

158/3866 13/423

1.00 0.36 (0.20–0.66)

73/2047 98/2242

1.00 1.47 (1.08–2.01)

Wald test P value

FDR

0.0202

0.0405

0.0073

0.0291

0.0396

0.1979

0.0037

0.0801

0.0065

0.0801

0.0356

0.2597

0.0053

0.0801

0.0210

0.1787

0.1174

0.4554

0.0429

0.2732

0.0036

0.0801

0.0079

0.0801

0.0368

0.1840

0.0251

0.1503

0.0218

0.1503

0.0130

0.0652

0.0366

0.4670

0.0009

0.0481

0.0157

0.2669

IFN-signaling pathway

Table VI. Continued

IRF2 (rs807684) AA/AG GG Summary score (0–2) (4–4) (6–6) (8–10) Ptrend

Death/person-years

HRR (95% CI)

164/3944 7/346

1.00 0.30 (0.14–0.66)

11/558 41/984 94/2227 25/521

1.00 2.68 (1.36–5.31) 3.32 (1.75–6.29) 4.85 (2.34–10.05) ,0.0001

Supplementary material The Supplementary Table can be found at http://carcin.oxfordjournals. org/ Funding National Cancer Institute (CA48998 and CA61757). This research also was supported by the Utah Cancer Registry, which is funded by Contract #N01-PC-67000 from the NCI, with additional support from the State of Utah Department of Health, the Northern California Cancer Registry and the Sacramento Tumor Registry. Acknowledgements The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official view of the NCI. We would like to acknowledge the contributions of Dr Bette Caan, Donna Schaffer and Judy Morse at the Kaiser Permanente Medical Care Program in Oakland, California; Jennifer Herrick, Sandra Edwards, Roger Edwards and Leslie Palmer at the University of Utah and Drs Kristin Anderson and John Potter at the University of Minnesota for data management and collection. Conflict of Interest Statement: None declared.

References 1. Gough,D.J. et al. (2008) IFNgamma signaling-does it mean JAK-STAT? Cytokine Growth Factor Rev., 19, 383–394. 2. Harada,H. et al. (1998) The role of interferon regulatory factors in the interferon system and cell growth control. Biochimie, 80, 641–650. 3. Saha,B. et al. (2010) Gene modulation and immunoregulatory roles of interferon gamma. Cytokine, 50, 1–14. 4. Platanias,L.C. et al. (1999) Signaling pathways activated by interferons. Exp. Hematol., 27, 1583–1592. 5. Bach,E.A. et al. (1997) The IFN gamma receptor: a paradigm for cytokine receptor signaling. Annu. Rev. Immunol., 15, 563–591. 6. Taniguchi,T. et al. (2001) IRF family of transcription factors as regulators of host defense. Annu. Rev. Immunol., 19, 623–655. 7. Schuhmann,D. et al. (2011) Interfering with interferon-gamma signalling in intestinal epithelial cells: selective inhibition of apoptosis-maintained secretion of anti-inflammatory interleukin-18 binding protein. Clin. Exp. Immunol., 163, 65–76. 8. Talseth,B.A. et al. (2007) Lack of association between genetic polymorphisms in cytokine genes and disease expression in patients with hereditary non-polyposis colorectal cancer. Scand. J. Gastroenterol., 42, 628–632. 9. Lee,Y.S. et al. (2010) Association between interleukin-4R and TGF-beta1 gene polymorphisms and the risk of colorectal cancer in a Korean population. Colorectal Dis., 12, 1208–1212. 10. Curtin,K. et al. (2010) Exploring multilocus associations of inflammation genes and colorectal cancer risk using hapConstructor. BMC Med. Genet., 10, 170–181. 11. Slattery,M.L. et al. (1997) Energy balance and colon cancer—beyond physical activity. Cancer Res., 57, 75–80. 12. Slattery,M.L. et al. (2003) Physical activity and colorectal cancer. Am. J. Epidemiol., 158, 214–224. 13. Edwards,S. et al. (1994) Objective system for interviewer performance evaluation for use in epidemiologic studies. Am. J. Epidemiol., 140, 1020–1028.

Wald test P value

FDR

0.0025

0.0639

14. Samowitz,W.S. et al. (2002) Prognostic significance of p53 mutations in colon cancer at the population level. Int. J. Cancer, 99, 597–602. 15. Slattery,M.L. et al. (2000) Associations between cigarette smoking, lifestyle factors, and microsatellite instability in colon tumors. J. Natl Cancer Inst., 92, 1831–1836. 16. Samowitz,W.S. et al. (2000) Relationship of Ki-ras mutations in colon cancers to tumor location, stage, and survival: a population-based study. Cancer Epidemiol. Biomarkers Prev., 9, 1193–1197. 17. Slattery,M.L. et al. (2007) Diet and lifestyle factor associations with CpG island methylator phenotype and BRAF mutations in colon cancer. Int. J. Cancer, 120, 656–663. 18. Conneely,K.N. et al. (2007) So many correlated tests, so little time! rapid adjustment of P values for multiple correlated tests. Am. J. Hum. Genet., 81, 1158–1168. 19. Slattery,M.L. et al. (2007) IL6 genotypes and colon and rectal cancer. Cancer Causes Control, 18, 1095–1105. 20. Slattery,M.L. et al. (2004) Dietary calcium, vitamin D, VDR genotypes and colorectal cancer. Int. J. Cancer, 111, 750–756. 21. Caan,B.J. et al. (1998) Body size and the risk of colon cancer in a large case-control study. Int. J. Obes. Relat. Metab. Disord., 22, 178–184. 22. Slattery,M.L. et al. (2011) Genetic variation in the TGF{beta}-signaling pathway and colon and rectal cancer risk. Cancer Epidemiol. Biomarkers Prev., 20, 57–69. 23. Slattery,M.L. et al. (2005) Energy balance, insulin-related genes and risk of colon and rectal cancer. Int. J. Cancer, 115, 148–154. 24. Slattery,M.L. et al. (2003) Energy balance and rectal cancer: an evaluation of energy intake, energy expenditure, and body mass index. Nutr. Cancer, 46, 166–171. 25. McGillicuddy,F.C. et al. (2009) Interferon gamma attenuates insulin signaling, lipid storage, and differentiation in human adipocytes via activation of the JAK/STAT pathway. J. Biol. Chem., 284, 31936– 31944. 26. Ahnen,D.J. (1998) Colon cancer prevention by NSAIDs: what is the mechanism of action? Eur. J. Surg. Suppl., 582, 111–114. 27. Freedman,A.N. et al. (2009) Colorectal cancer risk prediction tool for white men and women without known susceptibility. J. Clin. Oncol., 27, 686–693. 28. Kune,G.A. (2000) Colorectal cancer chemoprevention: aspirin, other NSAID and COX-2 inhibitors. Aust. N. Z. J. Surg., 70, 452–455. 29. Sandler,R.S. (1996) Aspirin and other nonsteroidal anti-inflammatory agents in the prevention of colorectal cancer. Important Adv. Oncol., 123–137. 30. Kode,A. et al. (2006) Differential effects of cigarette smoke on oxidative stress and proinflammatory cytokine release in primary human airway epithelial cells and in a variety of transformed alveolar epithelial cells. Respir. Res., 7, 132. 31. Ratovitski,E.A. (2011) DeltaNp63alpha/IRF6 interplay activates NOS2 transcription and induces autophagy upon tobacco exposure. Arch. Biochem. Biophys., 506, 208–215. 32. Hussain,S.P. et al. (2007) Inflammation and cancer: an ancient link with novel potentials. Int. J. Cancer, 121, 2373–2380. 33. Bouker,K.B. et al. (2005) Interferon regulatory factor-1 (IRF-1) exhibits tumor suppressor activities in breast cancer associated with caspase activation and induction of apoptosis. Carcinogenesis, 26, 1527–1535. 34. Tekautz,T.M. et al. (2006) Evaluation of IFN-gamma effects on apoptosis and gene expression in neuroblastoma—preclinical studies. Biochim. Biophys. Acta, 1763, 1000–1010. 35. Liu,F. et al. (2011) TNFalpha cooperates with IFN-gamma to repress BclxL expression to sensitize metastatic colon carcinoma cells to TRAILmediated apoptosis. PloS One, 6, e16241. Received April 11, 2011; revised July 27, 2011; accepted August 8, 2011

1667